hexagon logo
Register Sign in
  • State Not Answered
  • Replies 7 replies
  • Subscribers 11 subscribers
  • Views 167 views
  • Users 0 members are here
Options
Related

Static equilibrium

Hamed Hani
Hello Adams Community,
 
I am using Adams MaxFlex to run a nonlinear simulation.
 
I am generating the bulk data file in my code and importing the resulting `BDF` file into my model.
 
My model runs perfectly fine when I use a coarse mesh size like this image(the diameter of the circle is 4 cm):
5
But, as soon as I try a finer mesh size (dividing by 2) like this image, I get static equilibrium simulation convergence error(I don't change anything else about the model).
10
I tried to tweak the solver settings as it is suggested in this article:
 
 
I tightened and loosened the limits on translations (TLIMIT) and rotations (ALIMIT) while I relaxed the error, but had no success getting the simulation to converge. I increased the MAXIT value for convergence, but still no success.
 
Does anyone have any idea what might cause this problem?
 
solver
  • 0
    try searching 'equilibrium' in SimCompanion. you will quickly find out you are missing STABILITY and IMBALANCE to play with. Read about these and use with care equilibrium is an art
  • 0
    Thanks for response Alexis!
     
    I did follow the article on troubleshooting static equilibrium simulations. I tried to set the IMBALANCE to 1N and played with TLIMIT and ALIMIT, but I could not reach equilibrium. Do you know any other reasons for failing to converge in static simulation besides what's already discussed in the article?
     
     
  • 0
    Hi,
    you may want to play way more with the Imbalance, Stability... a single try with the Imbalance is likely not to work. check equilibrium in SimCompanion and read a few articles. practice will make you succeed
    the error/delta in your image are over the roof, they point to motion_r3 which may be at fault, maybe not.
    this is normally user error, make sure there are no lose parts, otherwise these will continuously fall in space (if you use gravity) and never reach equilibrium. Try running a very short dyn (0.1-0.5s) simulation with a non coarse time step (~0.01) and find out if something moves out of rance.
  • 0
    Hello,
     
    Thank you very much! I set the imbalance to a very large value. The static simulation failed to converge, but the dynamic simulation ran successfully. I see a very large force value at the initial time step, but the rest of the simulation appears to be fine. Does it matter to reach a static equilibrium before running the dynamic simulation?
     
     
  • 0
    setting the imbalance to a very lage value makes no sense you are artificially modifying a setting without understanding what it is about. Yo need to play with all the settings not just one.
    the very large force at initialisation is key to tell you there is something not right in the model
    Yes it matters a lot to set static equil first, same as in real life
  • 0
    How about trying the good old funneling technique ?
    I.e. a consecutive row of sim/sta commands starting with "loose" parameters getting "stiffer" ?
     
    Like:
    EQUILIBRIUM/error=10,stability=5,imbalance=10
    sim/stat
    EQUILIBRIUM/error=1,stability=1,imbalance=1
    sim/stat
    EQUILIBRIUM/error=0.1,stability=0.1,imbalance=0.1
    sim/stat
    EQUILIBRIUM/error=0.01,stability=0.01,imbalance=0.01
    sim/stat
     
    Generally speaking ADAMS static solver isn't what I'd call perfect or satisfying all customers needs. It's more like a capricious diva.
    Especially it doesn't like things creating sharp gradients over a short distance like contacts or flex-bodies or worst of all contact between flex bodies.
    In this case it may help changing the length units.
     
    Whenever I had models that still don't find equillibrium, I gave a sh... and just cut off the first few seconds of the analysis if possible.
     
  • 0
    Hi Martin,
    Thanks for the answer!
     
    I tried so many combination of error, stability, and imbalance and I could not reach equilibrium at all. What I ended up doing was to run a 0.5 second simulation of the model without the flexible part to reach equilibrium. Then, I copied the resulting model and added the flexible part. I successfully ran the whole simulation (with the flexible part) without any equilibrium issues with this method.
     
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement